Antigenic Variation of Streptococcus mutans Colonizing

Transcription

1 INFECTION AND IMMUNITY, Dec. 1975, P Copyright C) 1975 American Society for Microbiology Vol. 12, No. 6 Printed in USA. Antigenic Variation of Streptococcus mutans Colonizing Gnotobiotic Rats D. BRATTHALLI* AND R. J. GIBBONS Forsyth Dental Center, Boston, Massachusetts Received for publication 18 July 1975 Strains of Streptococcus mutans representative of serotypes b and d exhibited antigenic variation in both the oral cavity and in the intestinal canal of gnotobiotic rats. Laboratory-maintained cultures did not vary. The antigenic alterations observed were: (i) loss of detectable levels of both weakly reacting "strain" antigens and the type antigen; (ii) decreased production of the type antigen; (iii) production of altered type antigen; and (iv) production of an antigen not possessed by the parent strain. Immunization of animals before monoinfection with S. mutans strain Bob-1 (serotype d) appeared to increase the rate of emergence of antigenically altered mutants in the intestinal canal, and more diversely altered isolates were obtained. Antigenic variation may account in part for the variation noted by several investigators in attempting to immunize animals against S. mutans-induced dental caries. human teeth (24). Since antibodies, including those of salivary IgA, can inhibit the attachment of bacteria to oral surfaces (28), their presence in oral secretions should influence the colonization of affected bacteria. The agglutinating activities of salivery IgA preparations reactive against strains of oral streptococci have recently been found to change over a 150- day period, suggesting that the populations of streptococci colonizing the mouth also fluctuate over time (5). The present report describes changes in the antigenic composition of populations of Streptococcus mutans, an organism associated with dental caries activity, while colonizing the oral cavity and intestinal canal of gnotobiotic rats. Several infectious agents have been shown to exhibit antigenic variation during the course of colonizing a mammalian host. This phenomenon, sometimes referred to as antigenic drift, has been classically associated with influenza virus infections, and it probably occurs with other viral agents (9). Antigenic variation occurs in protozoan (6) and bacterial infections as well. In the case of bacteria, the serotypes of Escherichia coli inhabiting the human intestinal canal have been observed to change over time (7, 19). Similarly, populations of Vibrio cholerae in monoinfected gnotobiotic rats change serotype, probably because antibodies induced to certain antigenic components of the initial colonizing organisms cause the selection of antigenically altered mutants (17, 20). Popu- 'lations of Campylobacter fetus also have been MATERIALS AND METHODS reported to exhibit antigenic variation while Cultures and cultural conditions. S. mutans causing persistent venereal vibriosis in cows strain FA-1 (serotype b) was a rat isolate kindly (8), and it has been suggested that this enables provided by R. J. Fitzgerald. S. mutans strain Bob-1 the organism to elude the host's immune response (1, 8). In streptococci, it has been shown plaque (kindly provided by R. Berkowitz). The orga- (serotype d) was freshly isolated from human dental that both group and type antigens can be lost nisms were maintained by weekly transfer in Trypticase soy broth during serial mouse passages (14, 29). (BBL) and on mitis-salivarius agar plates (Difco). Human oral sections frequently contain Experimental procedures. A group of four germfree Sprague-Dawley rats, 26 days old, was main- detectable titers of antibodies reactive with strains of bacteria that normally inhabit the tained in flexible plastic isolators as previously described (10) and fed diet 2000 (12). The animals were mouth (5, 13, 22, 28). In addition, many bacteria present in saliva have been found to be infected with S. mutans strain FA-1, and oral swabbings and fresh fecal pellets were collected from coated with immunoglobulin (Ig) A (3), and significant quantities of both IgA and IgG are each animal periodically up to 94 days after infection. The samples were streaked on mitis-salivarius present in the microbial deposits that collect on agar plates, which were incubated in Brewer jars filled with 10% CO2, 10% H2, and 80% N2 at 35 C. I Present address: Faculty of Odontology, Fack, S Random colonies were picked from plates of each Goiteborg 33, Sweden. sample; these were restreaked to obtain pure cul- 1231

2 1232 BRATTHALL AND GIBBONS tures. Antigen extracts of each isolate were assayed by immunodiffusion for the presence of the b and n antigens previously described in the parent strain of FA-1 (4). Control antigen extracts were also prepared from 40 isolates derived from the original culture used to infect the animals. In a second experimental series, two groups of four germfree rats were housed in separate isolators. One group was immunized with 0.1 ml of a vaccine of S. mutans strain Bob-1 in each footpad; the injections were repeated 1 week later. The vaccine used was prepared by mixing equal quantities of a saline suspension containing 109 formalinkilled, Trypticase soy-grown streptococci with Freund incomplete adjuvant. Both the immunized and nontreated control animals were infected with the same culture of S. mutans strain Bob-1 1 week after the last injection. Oral swabbings and freshly collected fecal pellets were obtained periodically, and random isolates were obtained as described above. Antigen extracts of each isolate were analyzed for the presence of the d and s antigens of strain Bob-1 by immunodiffusion. Serological methods. Overnight Trypticase soy broth cultures of each isolate were used for preparation of the antigen extracts. The streptococci were harvested by centrifugation, washed, and suspended in saline. Each suspension was adjusted on the basis of optical density to contain 3.0 x 10, to 3.5 x 108 streptococci/ml. A 40-ml sample of each suspension was used for antigen extraction according to the method described by Rantz and Randall (18). All extracts for a given sampling period were made and autoclaved simultaneously to minimize variation in their preparation. Rabbit antiserum was prepared by using heatkilled cells of S. mutans according to an immunization schedule described previously (4). Cells used for immunization were grown in a sucrose-free, predialyzed Trypticase yeast extract medium. Antiserum to S. mutans strain FA-1 was used to analyze antigen extracts of isolates derived from animals infected with this organism, whereas an antiserum prepared to S. mutans strain B13 (serotype d) was used to analyze isolates obtained from animals infected with strain Bob-1. Immunodiffusion was carried out in 1.2% agar gels, using the microplate technique described by Wadsworth (26) for the FA-1 antigens or the Gelman instrument (Ann Arbor, Mich.) gel diffusion apparatus for antigens of Bob-1 isolates. RESULTS Antigenic variation in populations of S. mutans strain FA-1 (serotype b). The anti- FA-1 serum uniformly produced two distinct precipitin bands with antigen extracts of all 40 isolates derived from the culture used for infecting the rats (Fig. 1). The dense band represents the b type antigen and the weak band represents the n antigen as previously described (4). All isolates obtained from rats 5 to 10 days after monoinfection with S. mutans strain FA-1 contained b antigen, but the n antigen was not INFECT. IMMUN. detected in the extract of 1 of 11 fecal isolates (Table 1). The percentage of isolates lacking detectable n antigen increased over time, and 94 days after initial infection, 61% of the fecal isolates and 21% of the oral isolates no longer possessed detectable levels of n antigen under the standardized conditions of extraction used (Table 1; Fig. 2). In addition, one oral isolate that no longer possessed detectable levels of either the b or n antigen was found on day 62 (Fig. 2). Other isolates altered with respect to the b antigen were cultured from additional animals at subsequent sampling periods (Table 1). By the end of the experimental period, isolates altered in either the n or b antigen were isolated from every animal studied (Table 1). Electron microscopic examination (kindly performed by Z. Skobe, Forsyth Dental Center) of strains no longer producing detectable levels of either the b or n antigen, or both, did not reveal differences in their cell surface morphology that could be correlated with the antigenic changes. Antigenic variation in populations of S. mutans strain Bob-I (serotype d). Antiserum to S. mutans strain B13 (type d) produced a dense precipitin band corresponding to the d antigen and a weaker band, designated s antigen, with extracts derived from isolates from the parent strain Bob-1 culture (Fig. 3). Isolates from untreated animals were similarly homogeneous with respect to the antigenic components monitored 4 days after infection, but isolates that no longer exhibited detectable levels of s antigen were obtained after 14 days (Table 2; Fig. 4). Comparably altered isolates were present in all subsequent sampling periods, but no isolates that were altered with respect to the d antigen were obtained from these animals during the experimental period. The populations of S. mutans strain Bob-1 colonizing previously immunized animals appeared to exhibit greater antigenic variation; this was particularly evident among fecal isolates (Table 3). Forty-one days after infection, 80% of the isolates no longer possessed detectable levels of s antigen, and a few oral and fecal isolates exhibited weak d antigen reactivity (Fig. 5), though none were found which totally lacked detectable levels of this antigen. Extracts of other isolates were found to produce a spur with the parent d antigen (Fig. 6), indicating another type of antigenic alteration (Table 3). At day 53, two fecal isolates produced a precipitin band with the anti-b13 serum that was not detected in extracts of the parent strain Bob-1 (Fig. 6). This antigen has been referred to as "x." At the end of the experiments, 43%, of the total fecal isolates displayed this compo-

3 1 Fa- 1, parent stra in b an t i - Fa -1 2 Fa -1, re isolates 4 0 a n t - Fa Bobl, parent strain daot a nt i -B d -sa' d Bob l,reisolates; anti - B 13 in center basin Downloaded from Bob 1, parent strain 0 dm s (d) 0 0¾ /0 0 x Bob 1, d//o3d'~~~ re s o Ila t e d a n ti B1 3 Bob l,reisolates; anti- B13 in center basi n on October 16, 2017 by guest FIG. 1. Comparative double-diffusion in gel analyses of two antigen extracts of the parent strain Fa-i, which was used for infection of the animals. FIG. 2. Analyses oftwo antigen extracts ofreisolates from rats infected with Fa-i. One isolate shows lack of detectable amounts of n antigen, the other extract of n and b antigens. FIG. 3. Analyses of antigen extracts ofparent strain Bob-i. FIG. 4. Analyses of antigen extracts of three reisolates of strain Bob-i. One isolate shows the original composition (d +, s+), and one isolate contains the "new" antigen (d +, x +, s -). The last extract shows the d antigen only. FIG. 5. Antigen extracts of reisolates of strain Bob-i tested against anti-b13 serum. Note the weak precipitin band corresponding to the d antigen in extract in lower left basin. FIG. 6. Analyses of antigen extracts of a parent strain and reisolate of Bob-i. Spur formation in the d precipitin band indicates a partial change in the type antigen ofthe reisolate. The reisolate also contains the s and x antigens. 1233

4 1234 BRATTHALL AND GIBBONS TABLE 1. Antigenic variation of isolates obtained from gnotobiotic rats monoinfected with Streptococcus mutans strain FA-1 % of isolates that No. of antigenically altered isolates/no. of Total isolates were: Days Source of isolates examined altered/total inafter isolates isolates Deficient Deficient infection Rat 1 Rat 2 Rat 3 Rat 4 examined in b in n antigen antigen 5-10 Oral 0/3 0/3 0/3 0/3 0/ Fecal 1/5 0/2 0/2 0/2 1/ Oral 0/5 0/4 0/4 0/1 0/ Fecal 1/11 0/3 0/2 0/1 1/ Oral 0/5 1/5 0/5 0/5 1/ Fecal 0/5 0/5 1/5 0/5 1/ Oral 1/5 0/2 0/4 1/5 (1)a 2/16 (1) 6 13 Fecal 0/5 0/3 1/2 1/3 2/ Oral 1/4 2/4 (2) 0/4 1/4 (1) 4/16 (3) Fecal 1/6 2/6 2/6 1/6 6/ Oral 2/4 1/8 1/8 1/4 5/ Fecal 3/8 1/2 5/6 (2) 2/2 11/18 (2) a Isolates without detectable b antigen are in parentheses. TABLE 2. Antigenic variation of isolates obtained from gnotobiotic rats monoinfected with Streptococcus mutans strain Bob-i % of isolates that Day No. of antigenically altered isolates/no. of Total isolates were: after Source of isolates examined altered/total infection isolates isolates Deficient Deficient Rat 1 Rat 2 Rat 3 Rat 4 examined in d in s antigen antigen 4 Oral 0/5 0/5 0/5 0/5 0/ Fecal 0/5 0/5 0/5 0/5 0/ Oral 2/5 0/4 1/4 1/5 4/ Fecal 2/4 1/4 1/4 1/4 5/ Oral 1/1 2/5 2/4 1/5 6/ Fecal 4/5 2/5 1/4 1/5 8/ Oral 2/7 2/6 2/4 2/6 8/ Fecal 2/5 2/5 2/5 2/4 8/ Oral 0/5 0/5 0/5 0/5 0/ Fecal 0/5 1/5 0/5 2/4 3/ nent. None of the oral isolates from the immunized animals and none of the isolates from nonimmunized animals were found to contain detectable levels of x antigen. DISCUSSION The present study has shown that the populations of S. mutans colonizing the oral cavity and the intestinal canal of gnotobiotic rats change antigenic composition. Although the antisera used to monitor antigenic changes produced only two well-defined precipitin bands against antigen extracts of the two strains studied, a variety of antigenic alterations was observed when assayed by immunodiffusion. The most frequently observed alteration was the loss of detectable levels of the weakly reacting antigen of either S. mutans strain FA-1 (serotype b) or strain Bob-1 (serotype d); these are INFECT. IMMUN. probably more strain-specific antigens than the stronger-reacting type antigens used for taxonomic grouping. However, loss of detectable levels of the type antigen was noted in some isolates derived from strain FA-1, and in the case of strain Bob-i, isolates were obtained that exhibited either a decreased production of this component or an alteration such that it no longer exhibited patterns of identity with the type d antigen of the parent strain. In addition, several isolates from strain Bob-i-infected rats produced an antigenic component reactive with anti-strain B13 sera that was not evident in the antigen extract of the parent strain. It is possible that other new antigens or determinants would have been detected if antisera to additional strains had been used. The colonial morphology of several isolates differed from that of their parent strains on mitis-salivarius agar,

5 VOL. 12, 1975 ANTIGENIC VARIATION OF S. MUTANS 1235 TABLE 3. Antigenic variation of isolates obtained from gnotobiotic rats immunized and subsequently monoinfected with Streptococcus mutans strain Bob-i % of total isolates altered No. of antigenically altered isolates/no. Total isolates with respect to: Day Source of isolates examined altered/total Defi- Defi- Proinfection isolates Dsci- cie Pro- Rat 1 Rat 2 Rat 3 Rat 4 examined iency ciency duction imd ins of x antigen antigen antigen 4 Oral 0/5 0/5 0/5 0/5 0/ Fecal 0/5 0/5 0/5 0/5 0/ Oral 1/4 0/2 0/3 0/2 1/ Fecal 0/3 2a/4 1a/4 1/5 4a/16 Oa Oral 0/5 No sample 1"/1 1/8 2a/14 Oa 14 0 Fecal 6/6 6/9 5/6 4/5 21/ Oral 1/5 1/4 2/7 2/5 6/ Fecal 4/4 5/5 (2)" 2/2 5/5 16/16 (2) Oral 0/3 0/5 1/4 1/5 2/ Fecal 2/4 (1) 6/6 (4) 5/6 (4) 6/7 (1) 19/23 (10) a Some isolates possessed low levels of d antigen. bisolates with x antigen are in parentheses. but these differences did not correlate directly with the antigenic components monitored. These morphological changes probably reflect other alterations in the surface components of the S. mutans strains studied that were nonreactive with the antisera used. It is apparent from these observations that the antigenic composition of the populations of human and rodent strains ofs. mutans can vary widely during the course of colonizing experimental animals. The importance of antibodies as a mechanism for the selection of altered antigenic types has been demonstrated by Willers and Alderkamp in vitro (27). By subcultivating a nonhemolytic streptococcal strain several times in medium containing antiserum against the type antigen, they succeeded in isolating organisms lacking detectable quantities of this antigen. Studies of antigenic variation in V. cholerae colonizing gnotobiotic mice (17, 20) and of C. fetus infecting cattle (8) have suggested that the microbial population changes were mediated by an immune response to the initial colonizing populations that favored the selection of antigenically altered mutants. The observation that prior immunization increased the rate of emergence of antigenically altered mutants of S. mutans in the intestinal canal of monoinfected rats is consistent with this possibility; moreover, more diversely altered isolates were obtained from artificially immunized rats. The population changes observed are also consistent with the finding that the agglutinating activities of human salivary IgA preparations reactive with strains of indigenous streptococci, including S. mutans and S. salivarius, vary over time (5). In addition, strains of S. salivarius also exhibit antigenic variation while colonizing gnotobiotic rats (T. H. Howell and R. J. Gibbons, Abstr. Gen. Sess. Am. Assoc. Dent. Res. 1975, abstr. 628), and hence antigenic drift appears to be common among bacteria that colonize a mammalian host. Since strain FA-1 has been maintained in the laboratory for over 10 years without displaying evidence of antigenic alterations, and because isolates derived from cultures of the parent strains consistently yielded identical precipitin patterns, it is clear that the experimental animals created a selection pressure favoring the emergence of antigenically altered variants that is not present during in vitro cultivation. S. mutans has been associated with dental caries activity in both humans and experimental animals (11, 21), and several studies in recent years have attempted to induce immunity to tooth decay by using vaccines of S. mutans cells. Thus, hamsters (A. Gaffar, H. W. Marcussen, J. Ruffner, and R. C. Kestenbaum, Abstr. 48th Gen. Meet. Int. Assoc. Dent. Res. 1970, abstr. 304)., conventional and gnotobiotic rats (2, 16, 25), and two species of subhuman primates (2, 15) have been immunized by systemic administration of vaccines of S. mutans, and reductions in dental decay have varied between 0 and 60 to 80% after infection. In addition, several investigators have experienced difficulties in reproducing experiments (23, 25), and differences in protection have been noted between individual animals, particularly with primates (15). The marked variation in antigenic composition of S. mutans observed in the present study, especially in animals that had been systemically immunized, probably accounts for some of these variable experiences. Based upon the observations made, it seems likely that the

6 1236 BRATTHALL AND GIBBONS populations of S. mutans colonizing experimental animals for periods of weeks or months after immunization only partially resemble the antigenic composition of the parent strain used for vaccine production. Therefore, the probability of antigenic drift occurring in S. mutans populations must be considered in attempts to reduce dental caries by immunization procedures. To date, most attempts to induce immunity to dental caries have used animals that were heavily infected with S. mutans (2, 15, 16, 23, 25). However, it would appear that once this organism attains high population levels, the probability of antigenically altered mutants being present is also high, and hence antigenic drift likely occurs. Therefore, it appears that artificially induced immunity may be more promising in preventing initial colonization of S. mutans rather than in eliminating this organism after it has established an infection. ACKNOWLEDGMENTS This investigation was supported in part by Public Health Service grant DE from the National Institute of Dental Research and in part by a grant from the Colgate- Palmolive Co. Susan Edelstein is thanked for excellent technical assistance. LITERATURE CITED 1. Beale, G. H., and J. F. Wilkinson Antigenic variation in unicellular organisms. Annu. Rev. Microbiol. 15: Bowen, W. H A vaccine against dental caries. A pilot experiment in monkeys. Br. Dent. J. 126: Brandtzaeg, P., I. Fjellander, and S. T. Gjeruldsen Adsorption of immunoglobulin A onto bacteria in vivo. J. Bacteriol. 96: Bratthall, D Immunodiffusion studies on the serological specificity of streptococci resembling Streptococcus mutans. Odontol. Revy 20: Bratthall, D., and R. J. Gibbons Changing agglutination activities of salivary immunoglobulin A preparations against oral streptococci. Infect. Immun. 11: Brown, I. N., K. N. Brown, and L. A. Hills Immunity to malaria: the antibody response to antigenic variation by Plasmodium knowlesi. Immunology 14: Cooke, E. M., S. Ewins, and R. A. Shooter Changing faecal populations of Escherichia coli in hospital medical patients. Br. Med. J. 4: Corbeil, L. B., G. G. D. Schurig, P. J. Bier, and A. J. Winter Bovine veneral vibriosis: antigenic variation of the bacterium during infection. Infect. Immun. 11: Fenner, F., B. R. McAnslan, C. A. Mims, J. Sambrook, and D. 0. White The biology of animal viruses, 2nd ed., p Academic Press Inc., New York. 10. Gibbons, R. J., S. S. Socransky, and B. Kapsimalis. INFECT. IMMUN Establishment of human indigenous bacteria in germ-free mice. J. Bacteriol. 88: Gibbons, R. J., and J. van Houte Dentel caries. Annu. Rev. Med. 26: Keyes, P. H., and H. V. Jordan Periodonal lesions in the Syrian hamster. III. Findings related to an infectious and transmissible component. Arch. Oral Biol. 9: Kraus, F. W., and J. Konno The salivary secretion of antibody. Ala. J. Med. Sci. 2: Lancefield, R., and G. Perlmann Preparation and properties of a protein (R antigen) occurring in streptococci of group A, type 28 and in certain streptococci of other serological groups. J. Exp. Med. 96: Lehner, T., S. J. Challacombe, and J. Caldwell An immunological investigation into the prevention of caries in deciduous teeth of Rhesus monkeys. Arch. Oral Biol. 20: McGhee, J. R., S. M. Michalek, J. Webb, J. M. Navia, A. F. R. Rahman, and D. W. Legler Effective immunity to dental caries: protection of gnotobiotic rats by local immunization with Streptococcus mutans. J. Immunol. 114: Miller, C. E., K. H. Wong, J. C. Feeley, and M. E. Forbines Immunological conversion of Vibrio cholerae in gnotobiotic mice. Infect. Immun. 6: Rantz, L. A., and E. Randall Use of autoclaved extracts of hemolytic streptococci for serological grouping. Stanford Med. Bull. 13: Robinet, H. G Relationship of host antibody to fluctuations of Escherichia coli serotypes in the human intestine. J. Bacteriol. 84: Sack, R. B., and C. E. Miller Progressive changes of vibrio serotypes in germfree mice infected with Vibrio cholerae. J. Bacteriol. 99: Scherp, H. W Dental caries: prospects for prevention. Science 173: Sirisinha, S Reactions of human salivary immunoglobulins with indigenous bacteria. Arch. Oral Biol. 15: Tanzer, J. M., G. J. Hageage, and R. H. Larson Variable experiences in immunization of rats against Streptococcus mutans-associated dental caries. Arch. Oral Biol. 18: Taubman, M. A Immunoglobulins of human dental plaque. Arch. Oral Biol. 19: Taubman, M. A., and D. J. Smith Effects of local immunization with Streptococcus mutans on induction of salivary immunoglobulin A antibody and experimental dental caries in rats. Infect. Immun. 9: Wadsworth, C A microplate technique employing a gel chamber compared with other micro and macroplate techniques for immune diffusion. Int. Arch. Allergy 21: Willers, J. M., and G. H. Alderkamp Loss of type antigen in a type III Streptococcus and identification of the determinant disaccharide ofthe remaining antigen. J. Gen. Microbiol. 49: Williams, R. C., and R. J. Gibbons Inhibition of bacterial adherence by secretory immunoglobulin A: a mechanism of antigen disposal. Science 177: Wilson, A. T Loss of group carbohydrate during mouse passages of a group A hemolytic streptococcus. J. Exp. Med. 81: